Oral health effects of non‑combustible nicotine products: a systematic review and network meta‑analysis of randomized controlled trials
- Alan Gor
- Jun 13
- 2 min read
13 June 2025
Giusy Rita Maria La Rosa a, Cinzia Del Giovane b c, Silvia Minozzi d, Jan Kowalski e, Iain Chapple f, Amaliya Amaliya g, Konstantinos Farsalinos h, Riccardo Polosa a i
Abstract
Objectives
To evaluate the oral adverse effects of non-combustible nicotine products (NCNPs) compared with each other, placebo, standard of care, no treatment and combustible cigarettes through a systematic review and network meta-analysis.
Data sources and study selection
Randomized controlled trials involving adult smokers and reporting oral adverse events (e.g., mouth irritation, dry mouth, aphthous ulcers) were included. PubMed, Scopus, and Cochrane CENTRAL were searched up to August 2024. Risk of bias was assessed using RoB 2, and evidence certainty with CINeMA.
Results
Thirty-six trials were included, with 21 contributing to the network meta-analysis. Most comparisons with placebo showed no significant differences across four primary outcomes. The odds of developing aphthous ulcers were significantly higher in the nicotine replacement therapy (NRT) gum group compared with standard of care (OR = 2.36; 95% CI: 1.05–5.30). Higher odds of mouth irritation were also observed for e-cigs (OR = 4.06; 95% CI: 1.67–9.85), NRT mouth spray (OR = 4.36; 95% CI: 1.14–16.63), NRT gum (OR = 4.25; 95% CI: 1.51–11.94) and snus (OR = 13.56; 95% CI: 1.07–171.52) when compared with standard of care. Sensitivity analyses confirmed the main findings. Secondary outcomes revealed isolated associations but were based on limited data. Evidence certainty was low to very low due mainly to imprecision and risk of bias.
Conclusions
NCNPs appear to be generally well tolerated. Most placebo comparisons showed no increased risk, although some products exhibited higher odds of aphthous ulcers and mouth irritation compared with standard of care. Better reporting of oral adverse events in RCTs is needed.
